Current antibody display technologies (phage, yeast, ribosome, mammalian, etc.) are limited because the quality of the selected antibody candidates is limited by the starting library from which they are generated. Approaches, such as combinatorial and “intelligent” antibody design approaches and hybridoma discovery approaches, often yield synthetic antibodies that present downstream complications including large scale expression difficulties, high risk of immunogenicity in patients, and lack of sufficient immune function other than high binding affinities. Few antibodies derived from display technologies have successfully passed clinical trials in the last decade, even when demonstrating positive pre-clinical characteristics. Currently, the ability to predict or understand the mechanism by which a particular antibody sequence recognizes and activates the immune response against a foreign target has remained elusive. Thus, there is a need in the art for methods to discover and generate antibodies that have high binding affinities, can be generated on a large scale, and have sufficient immune function. The methods described herein aim to utilize the millions of years of immune repertoire evolution to meet these needs and to further the understanding of these concepts and how they relate to the generation of antibodies. The methods described herein can be used to produce a library of antibody sequences and/or antibodies for selection of high quality antibody candidates.
The human antibody repertoire is almost unlimited in its complexity and size. As a result, combinatorial libraries have statistically been demonstrated to rarely yield correct heavy (VH) or light (VL) chain pairing. Others have focused on shuffling the only of the most frequently expressed framework families of complementarity determining regions (CDRs) (such as V3-23, V1-69, or matching VH and VL frequencies), and therefore limited repertoire diversity to a manageable size. It was expected that the most frequently expressed family would be more frequently selected and evolved during an immune response. Surprisingly, through the use of immune sequencing of human antibody repertoires, it has been discovered that there is no relation between antibody framework expression frequencies and the activation potential of an antibody in response to an immune challenge. The methods described herein can be used to design and/or generate a non-limiting antibody library to overcome these challenges for antibody discovery and selection. Autoimmune, cancer, infectious and normal/healthy donor libraries can be generated for personalized medicine to address fundamental unmet biological needs.